Nature Inspired Partitioning Clustering Algorithms: A Review and Analysis

Clustering algorithms are developed as a powerful tool to analyze the massive amount of data which are produced by modern applications. The main goal of these algorithms is to classify the data in clusters of objects, so that data in each cluster is similar based on specific criteria and data from two different clusters be different as much as possible. One of the most commonly used clustering methods is partitioning clustering method. So far various partitioning clustering algorithms are provided by researchers, among them inspiring the nature algorithms are the most popular used algorithms. In this paper some partitioning clustering algorithms inspiring by nature are described, and then these algorithms are compared and evaluated based on several standards such as time complexity, stability and also in terms of clustering accuracy on real and synthetic data sets. Simulation results have shown that combinational methods have good influence to increase the efficiency of algorithms and also the use of different operators can maintain population diversity and cause to reach a good answer in a reasonable time.